mlx-whisper

Transcribe audio files locally on Apple Silicon using MLX Whisper.

36|2|Updated Feb 9, 2026
One-click install
npx skills add https://github.com/malue-ai/dazee-small --skill mlx-whisper
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: mlx-whisper
Source: https://github.com/malue-ai/dazee-small/tree/main/skills/library/mlx-whisper
Command: npx skills add https://github.com/malue-ai/dazee-small --skill mlx-whisper

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires mlx-whisper, and includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill provides a significantly faster, privacy-preserving, and offline alternative to cloud-based speech-to-text services by leveraging Apple Silicon's MLX framework.

Core Features & Use Cases

  • Local Speech-to-Text: Transcribes audio files directly on your machine without internet connectivity.
  • High Speed: Achieves up to 10x faster transcription speeds compared to standard OpenAI Whisper.
  • Privacy Focused: Keeps all audio data and transcriptions local, ensuring data privacy.
  • Use Case: Quickly convert voice memos, meeting recordings, or any audio file into text for documentation, note-taking, or further processing, all while maintaining data security.

Quick Start

Use the mlx-whisper skill to transcribe the audio file named 'meeting_recording.mp3'.

Frequently Asked Questions about mlx-whisper

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I transcribe audio files locally on Apple Silicon without an internet connection?

You can perform local speech-to-text transcription directly on Apple Silicon by using the MLX Whisper model. This process operates entirely offline, ensuring audio data remains private while converting recordings into text.

Can I get word-level timestamps when converting speech to text using MLX?

Yes, speech-to-text conversion using the MLX Whisper model can output text with word-level timestamps. This allows for precise alignment of transcribed text with specific moments in the original audio file.

What is the best way to speed up speech-to-text transcription on Mac compared to standard OpenAI Whisper?

Using the MLX Whisper model on Apple Silicon achieves up to 10x faster transcription speeds compared to standard OpenAI Whisper. This local approach optimizes hardware acceleration to reduce processing time significantly.

Do I need an internet connection to use MLX for speech-to-text conversion?

No internet connection is required for speech-to-text transcription with MLX Whisper. The model runs completely offline on your local machine, guaranteeing that audio data and transcriptions never leave your device.

Does local speech-to-text on Apple Silicon support different model sizes for speed and accuracy trade-offs?

Yes, local speech-to-text using the MLX Whisper model supports various model sizes. You can select different models to balance transcription speed and accuracy based on your specific hardware and project requirements.

Why use offline transcription with MLX instead of cloud-based speech-to-text services?

Offline transcription with MLX prioritizes data privacy by keeping all audio local, avoiding cloud uploads. It leverages Apple Silicon hardware for high-speed processing, making it ideal for secure meeting recordings and voice memos.